- Installing
solaris - An introduction to
solaris - Pretrained models available in
solaris - Solaris API summary
- Solaris Tutorials and Cookbook
- Converting GeoJSON labels to COCO-formatted labels using
Solaris - Scoring model performance with the
solarispython API - Converting model outputs to vector format using the Python API
- Creating training masks with the
solarispython API - Tiling imagery and labels using the
solarisPython API - Training your own custom model using
solaris - Training included SpaceNet models with the
solarisPython API - Using the
solarisCLI to make training masks - Running a deep learning pipeline with the
solarisCLI - Using the
solarisCLI to score model performance - Creating reference CSVs for model training and inference
- Creating the YAML Configuration File
- Mapping vehicles with
solarisand thecowcdataset- Specify our directories for pre processing
- Initialize a tiling function
- Orgainze our data
- Tile our masks and convert them to GeoTiffs
- Dialate our masks to increase the size of our labels
- Calculate some basic statistics for z-scoring (normalizing) our imagery
- Hold out a city for testing
- Review some of our masks and images
- Create a csv file that lists our images and our masks for training and testing
- Edit your .yml file and begin training your model
- Time to inference
- Specify our directories for post-processing
- Post-processing- binarize our masks and convert them to polygons
- Check out our results
- Initialize a few more functions for scoring our results.
- Score our results
- Solaris Multimodal Preprocessing Library
- Solaris Multimodal Preprocessing Library
- Example 1: A Simple Pipeline
- Example 1 Follow-Up: Building a Reusable Class
- Solaris Multimodal Preprocessing Library
- Example 3 Follow-Up: Masking and Multimodal Datasets
- Epilogue
- The command line interface
- The Python API
- Reference
- Index
- Converting GeoJSON labels to COCO-formatted labels using